The processes of speciation have been matters of intense interest to evolutionary biologists for over a century, yet remain enigmatic. Because speciation is difficult to study as an active process, most work on the subject involves the comparison of minimally divergent taxa, presumed to have speciated only recently. Unfortunately, the morphological discontinuities used to delimit even closely related species may not develop until long after genetic radiation and reproductive isolation have evolved. It seems probable that many currently defined 'morpho-species' harbor cryptic complexes of sibling taxa, and if we are to understand speciation better, we must move the study of genetic radiation to the level of these vicariids. The purpose of this proposal is to develop statistical/genetic techniques to improve our taxonomic resolution below the 'morpho-species' level, and then evaluate these analytic tools on simulated data sets and on a set of carefully chosen biological test cases, where the existence of the sibling taxa can be inferred on other grounds. There are numerous genera whose taxonomic and evolutionary elucidation could profit from the improved resolution that should result from the techniques being developed here. In addition, many genera of agricultural of human health concern also harbor collections of troublesome sibling taxa, and the ability to separate individuals on simple genetic evidence could be of considerable utility for a variety of pest and disease vector control programs.